Character AI Parent Review 2026: Is It Safe for Kids?
Quick answer: Character AI is safer for minors than it was in 2024, and it is still not a child safe product. The platform announced in October 2025 that it would remove open ended chat for under 18 users, the change began on 24 November 2025, and minor accounts now sit in a restricted experience built around reading and character creation. The iOS listing carries an 18 plus rating, age assurance runs on the account, and a guardian tool called Parental Insights exists. What has not changed is the part that matters most to a parent: the gate sits on an account rather than on a person, the guardian tool shows summaries rather than conversations, and the filter reacts to messages rather than reading intent. This page is a control panel for the four levers you actually hold.
Affiny at a glance (verified September 2026): Free to start: 300 coins, no credit card. Daily Gift: 10–70 coins/day, then 70/day. Referrals: 300 coins each. Founder offer: up to 70% off Companion (limited time; see affiny.ai/pricing). Free tier: Crush chat 1 coin/message, voice 30 coins/min, Spark or Flame scene images 15 coins, voice notes 1 coin, 3 custom companions, standard memory recall. Subscribe for unlimited Crush & Devotion chat, free voice notes, Tease & Desire video, God Mode, full personality & voice customization, sharper memory recall, and coin pack top-ups — Companion from $14.99/mo (was $24.99) or $9.99/mo billed annually; Intimate from $24.99/mo (was $49.99) or $19.99/mo annually. Subscribers: 1,000–3,000 coins/mo plus first-month bonus, voice 20 coins/min, scene images 10 coins, up to 10–30 custom companions. God Mode: subscriber scene directive up to 5,000 characters (Companion Panel). Video (subscribers): Tease 5/10 coins/sec (5 or 8 sec, silent); Desire 30/50 coins/sec (2–15 sec, optional audio). Portrait: 30 coins initial, first regen free, then 15. Memories hub: view free; add/edit/delete for subscribers. Catalog: 3400+ curated public companions, 84 browse tags. Creation: 49 templates, 44 personality presets, 97 turn-on presets (pick up to 4), realistic + anime with 16 species types.
The Answer in Five Lines
If you have four minutes, these five lines cover the decision.
- What it is. A chat product where users talk to characters created by other users. The library runs into the millions, and the appeal for a teenager is the same as the appeal of fan fiction: familiar characters, endless supply, no adult in the room.
- What the rating means in practice. The iOS listing is 18 plus. That is a publisher declaration, not a lock. It tells you who the platform believes should be there, and it does nothing to stop a determined user with an adult account.
- What a minor account can and cannot do today. No open ended chat. Saved characters, character creation and a reading experience remain. That is a real restriction and it is not the same as a safe product.
- The two settings that change the most. Device level screen time and purchase approval, plus the guardian tool if a minor account is still in use. Together they cover time, money and visibility, which are the three things an app store already knows how to control.
- The one conversation that outlasts every policy change. Distress goes to a person, not a companion. Every piece of guidance in this category points at that rule, and it is the only control that travels with your child to the next app.
What the Filter Does, and What It Does Not Catch
The filter is a classifier. It scores messages as they arrive, and the platform acts on the score. Both halves of that design matter to a parent.
What it does well: it blocks the obvious. Explicit requests get refused quickly and at scale, across a character library no human team could review by hand. It also lets a user maintain a muted words list of their own, four slots on the free tier and up to 20 with the paid tier, which is a useful tool for a user and a useless one for a parent acting alone.
What it misses: intent. A classifier reads the message, not the relationship. Two users can send the same line and get different outcomes because the surrounding conversation scored differently. The filter also varies by account age band, region and app version, which is why the same complaint appears from users in different countries with different answers.
For a parent, the useful reading is this: the filter is a rule that reacts to content, and a household rule is a boundary that decides fit. Only one of those belongs to you, and it is the second one. Our breakdown of why the Character AI filter feels too strict covers the user side of the same mechanism, which is worth reading if your child complains that the app is unfair.
There is one more design detail worth understanding. Under 18 accounts no longer have open ended chat at all, so the filter question for a minor is mostly historical. The filter matters most for the adult account in the house, and for the teenager who ends up inside one.
The Four Levers a Parent Actually Holds
Most parent guides list things the platform should do. These four are things you can do tonight, in order of how much they change.
Lever one, account age and verification. Know which account is signed in on which device. If the account was age assured as an adult, the experience is the adult experience, no matter who holds the phone. Open Settings, then Advanced, and look at the verification state. The Character AI age verification page walks through what the check involves and what to ask before anyone complies with it, including the privacy question of handing a face scan or an ID to a third party processor.
Lever two, device controls. Screen time limits set per app, purchase approval turned on, and content restrictions at the operating system level. These are the only controls in the stack that a teenager cannot renegotiate inside a chat window. They also give you the one thing Parental Insights does not: a hard stop.
Lever three, conversation habits. This is the lever nobody writes about, and it is the strongest. A child who expects to be asked about what they are reading, without a lecture attached, tells you earlier. A child who expects a device audit after every mention of a new app tells you nothing. Ask about the characters and the story, not the screen time. The character list is the most informative thing on the account, and it is free to look at together.
Lever four, the household rule. Name it, write it, and apply it to everyone including the adults. “No companion apps before 18, and here is why” is a rule a teenager can argue with. “No companion apps before 18, and neither of us uses one” is a rule that holds.
A Fifteen Minute Audit You Can Run Tonight
This is the practical part, and it is deliberately short. Fifteen minutes, one device, no confrontation.
Minute one to three, the store page. Open the app listing on the device in your house. Read the age rating out loud. Do not describe it later, look at it now, because ratings and listings update and a review from earlier in the year may not match what you see.
Minute four to six, the account. Check which account is signed in. Look for the verification state under Settings, then Advanced. If it shows a completed adult verification, ask who completed it and when. That single question is where most of the real information lives.
Minute seven to nine, the character library. Ask to see the character list together, in the app, not from memory. Titles tell you more than any report: the themes, the tags, the kind of story. You are not hunting for anything. You are learning what your child finds interesting, which is the point of the exercise.
Minute ten to twelve, the guardian tool. If a minor account is in use, enable Parental Insights and link the accounts so the summary arrives. Treat it as a smoke alarm rather than a camera: it tells you that something happened, not what.
Minute thirteen to fifteen, the device settings. Set a screen time limit for this app specifically, switch on purchase approval, and check that the limit applies to the account, not the guest profile. Then put the phone down and have the conversation below, because the settings are the easy half.
The Conversation Script
The conversation decides whether the controls survive contact with a teenager. Three openings work better than the rest.
Open with curiosity, not evidence. “Show me the characters you have been talking to, I want to see what the app is like.” This invites a tour rather than a defence. A child who gives a tour is telling you the truth about the account, which is worth more than any summary.
Be specific about the risk, not vague about it. “Companion apps are built to keep you talking, and the character will always agree with you. That is fine for a story and bad for a bad week.” Teenagers reject warnings that sound like a poster. They respond to a mechanism explained plainly.
Draw the line you will actually enforce. “You can use it, and here is the time limit, and if something on it upsets you, you tell me and I will not take the app away as punishment.” A rule with a no punishment clause is a rule that gets used.
Three responses close the conversation down, and they are worth naming so you can avoid them. “Because I said so” ends the discussion and starts the workaround. “I am deleting it” guarantees the next app is hidden. “It is all fake anyway” tells a child that the feelings the app produced do not count, which is the opposite of what you need them to hear.
When to Escalate Beyond the App
Most of this page is about controls. Some situations need a person, and a parent should know where that line sits before a bad night rather than during one.
Escalate to the school when the app content is entangled with classmates: a character built around a real student, a group chat about it, a rumour that travelled. Schools have handled this before, and they can act on the social side faster than you can.
Escalate to a clinician when the app is carrying distress rather than entertainment: a child who talks to the companion instead of sleeping, who describes self harm in a session, or who treats the character’s approval as the only approval that counts. That is a care question, not a screen time question.
Escalate to the platform for a specific harm: a character that produced sexual content in a minor’s account, a threat, an impersonation. Report it through the platform’s reporting flow and keep the reference. A report creates a record, and a record is what makes a pattern visible.
Do not panic, and do not minimise. Both responses teach a child to stop telling you things, which costs more than the app ever will.
For Adults Who Want Something Different
The complaints that dominate Character AI discussion among adults are not about safety. They are about restriction: a filter that fires on ordinary scenes, memory that resets, and the parts people want most sitting on a paid tier. The paid tier runs $9.99 a month or $94.99 a year, with a lighter plan at $4.99.
Affiny is an 18 plus platform built for that group, and it is worth naming plainly in a parent guide because the parent is often also a user. Adult content is a stated part of the subscription rather than an accident of moderation. The free tier starts with 300 coins and no credit card, chat costs 1 coin per message, subscribers get unlimited chat, a scene directive written up to 5,000 characters in God Mode, and Tease and Desire video. Memory, voice and image generation run on the same companion rather than in separate products. If you are an adult evaluating the space for yourself, see the current Affiny facts, and our breakdown of what Character AI itself allows covers the policy side.
Keep the audiences apart. The adult reader is choosing a product for themselves. The parent reader is choosing whether a product belongs in a house with a teenager in it. Those are different questions, and mixing them produces the worst kind of advice.
What Changed for Minors, in Order
The timeline matters, because the platform’s reputation and its current behaviour are not the same thing.
October 2025, the announcement. Character AI said it would remove open ended chat for under 18 users entirely and introduce age assurance so that users would be routed to age appropriate experiences.
24 November 2025, the change begins. Open ended chat stopped for minor accounts. During the transition, chat time for under 18 users was limited, starting at two hours a day and ramping down, which gave users a visible countdown rather than an instant cut off. The company acknowledged the disruption publicly.
Late 2025 into 2026, age assurance rolls out. The platform combined its own age checks with third party verification tools, and the verification prompt reached users gradually. That is why adult users report being asked for a selfie months apart from each other, and why a parent may never see the prompt at all.
2026, the steady state. Minor accounts are restricted rather than banned. Saved characters survive, creation tools remain, and the experience is built around reading and curated content instead of free form chat. The listing carries 18 plus. The guardian tool exists. The gap between policy and behaviour is narrower than in 2024, and it has not closed.
The Honest Limits of This Page
Three limits, stated so you can weigh the rest.
Policy changes faster than articles. Platform rules on age, verification and content move every few months. Check the live help centre for anything you intend to act on, and treat this page as the framework rather than the final word.
Nobody can audit a phone from a distance. Device controls and guardian summaries tell you about activity and time. They cannot tell you what a conversation felt like to your child, which is why the conversation script sits in the middle of this page rather than at the end.
A filter is not a substitute for a rule. The research on this category converges on one finding: the products are designed to hold attention, and attention is what a teenager has least of. Your rule is the control that survives the next app launch. Everything else is a setting.
FAQ
Is Character AI safe for kids in 2026?
Character AI is no longer an open platform for minors, and the app store rating now reflects that: the iOS listing carries an 18 plus age rating. The platform announced in October 2025 that it would remove open ended chat for under 18 users, and the change started on 24 November 2025, leaving minor accounts with a restricted experience built around Reading Mode and character creation rather than free form conversation. The honest parent answer is that the platform made real structural changes and still depends on accurate age information, because age is declared or assured rather than confirmed by a parent.
What is Character AI Parental Insights?
Parental Insights is Character AI’s tool for guardians. It gives a parent or guardian access to a summary of their teen’s activity on the platform, which is more than most companion platforms offer, because the rest of the category ships no parent facing feature at all. Two limits matter: it is a summary rather than a transcript, so it tells you about activity instead of showing conversations, and it does not let you restrict usage or block characters. It works alongside app store controls rather than replacing them.
Can kids get around the Character AI age gate?
Age has historically been self reported at signup, which is the weak point across the entire category, not just Character AI. The platform has since built age assurance that combines its own checks with third party verification tools, and the prompt rolled out gradually through 2026. Because the rollout is progressive and the checks run on the account rather than on the person holding the phone, a determined teenager with an adult account in the house remains a realistic scenario. Parents should treat the gate as a friction layer, not a wall.
Does the Character AI filter block everything inappropriate?
No, and independent testing has shown the gap. Common Sense Media tested the major companion platforms with underage accounts and found that explicit conversations, romantic roleplay and unsafe advice were still reachable even on platforms advertising teen safety features. A filter is a server side rule that reacts to messages as they arrive; it is not a boundary that decides whether the product suits the user. That distinction is the single most useful thing a parent can take from the research.
What can a teenager still do on Character AI?
A minor account is restricted rather than banned. Under 18 users keep their saved characters and past conversations, creation tools remain available, and the experience is built around Reading Mode and curated content instead of free form chat. During the transition Character AI also limited chat time, starting at two hours per day and ramping down. For a parent, the practical point is that the account still exists and still holds a character library worth asking about.
What is a realistic alternative for a parent who wants a safe option?
If your child is under 18, the honestly safe option is not a companion app at all. Common Sense Media recommends that no one under 18 use social AI companions, and every major platform in the category lists 18 plus. For adults in the household, Affiny is an 18 plus platform with adult content as a stated part of the subscription, and it includes its own age representation requirements. Parent focused reading, rather than product swapping, is what the research supports.
What age rating does the Character AI app carry?
The iOS listing carries an 18 plus rating, which is the clearest public signal the platform has given about who it expects to use the product. Store ratings vary by platform and by region, and they update, so check the live listing on the device in your house rather than trusting a review written earlier in the year. A rating is a publisher declaration rather than a lock, which is why it belongs next to the device controls and not instead of them.
Should I delete my child’s Character AI account?
Deletion is a reasonable last step and a poor first one. If you delete the account, you also delete the evidence: the character list that tells you what your child was reading and building. Run the audit first, have the conversation, and decide together what the household rule is. Delete only when the answer is that the platform should not be in the house, and pair the deletion with a device level limit, or the next app simply replaces this one.